Page 1 of 2 12 Last
  • Jump to page:
    #1
  1. Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Oct 2004
    Posts
    121
    Rep Power
    15

    Sandboxes, sitewides, and SERPs... oh my!


    Okay. I'm pretty convinced that the sandbox is based purely on the age of the site according to Google (i.e. how long it has been indexed). I believe link age has to do with a totally seperate filter that is seperate from (and far less devastating than) the sandbox. Allow me to explain why I feel this way...

    I have this domain that I have as just a goof-off testing domain. This is a domain that was indexed in G several years ago, but was abandoned. It has been indexed in Google for a couple of months and has been #1 for a VERY non-competitve term. Now I know what you're thinking... "Anybody can come up for a very non-competitive term. Sandbox or not." I have no problem believing this. The fact is, I changed what keywords I'm targeting for this site now... and it's been in G long enough to where the "new site bonus" in the SERPs should be gone by now.

    My existing site (which I believe is sandboxed) is a store. This site has hundreds of pages. Let's call this site "Blue Widgets Store". A lot (but not all) of my back links has "Blue Widgets Store" as the anchor text. If I do a search for "Blue Widgets Store", my site doesn't show up at all... but it shows up #1 in the allin searches.

    Now... I've changed my test site to have "Blue Widgets Store" as the title. I've also put it in an H1 tag and in the alt text for an image. The only other thing on this page is a sentence saying "We've moved! Come visit The Blue Widgets Store at www.bluewidgetsstore.com". I then put up a sitewide link from my sandboxed site to this site with "Blue Widgets" as the anchor text. About three days later the test site is already in the SERPs for "Blue Widgets Store"... beating my original site that is still nowhere to be found. What kills me is... this site doesn't even have the word "Store" in any of the backlinks, and it's only being supported by a single sandboxed site! It's ranking better than its sole supporter for a keyword phrase the supporting site is aiming for! The supportee is only targeting a keyword subset.

    Isn't this kinda odd? Could it be because this test site was once indexed by Google years ago? I don't know what to think. I'm still an SEO newbie, so feel free to flame away.

    Comments on this post

    • dazzlindonna agrees : thanks for sharing this info
    • EGOL agrees : Thanks! Interesting. I wish more posts would offer this personal experience theme.
    • randfish agrees : Excellent information there, Olaf; thank you
    • earlpearl agrees : GREAT EXPERIMENT! Tell us more Olaf. This is one to watch.
  2. #2
  3. Contributing User
    SEO Chat High Scholar (3500 - 3999 posts)

    Join Date
    Mar 2003
    Location
    Louisiana, USA
    Posts
    3,874
    Rep Power
    24
    No reason for anyone to flame you. That's a great little experiment. I'd like to have a few more pieces of information. How old is the old site and how old is the new site? What is the PR for each site and how many backlinks (according to Yahoo) does each site have? And finally, about how many links constitutes the sitewide links that you put up on the new site to the old one? Thanks for sharing that!
  4. #3
  5. http://tinyurl.com/cz56g
    SEO Chat Mastermind (5000+ posts)

    Join Date
    Sep 2004
    Location
    D0RDRECHT NL
    Posts
    6,063
    Rep Power
    31
    Heh Olaf - you're on a roll. I can't even rep you right now

    That's a cool bit of science you posted there. If anything, it confirms that I should cherish my fairly old domain, and not let it go to waste because of any "evil" stuff or other - similar - temptations
  6. #4
  7. SEO Chat Skiller (1500 - 1999 posts)

    Join Date
    Jul 2004
    Location
    Seattle, WA
    Posts
    1,888
    Rep Power
    47
    Olaf - THat is fascinating insight into Google's operation of the sandbox. Clearly they recognize that your site "should be" ranking for the term and are even passing on its outbound link value, but they have penalized it such that it simply cannot rank for certain phrases.

    This lends a lot of credibility and evidence to the sandbox theories out there. Thank you for sharing.
    CEO & Co-founder of SEOmoz
  8. #5
  9. Free the SB
    SEO Chat Skiller (1500 - 1999 posts)

    Join Date
    May 2004
    Location
    DC region
    Posts
    1,832
    Rep Power
    20
    That is truly a great find.

    Olaf it would be helpful to keep this up and report on serps progress for the old site.
    255 chars s*ck
  10. #6
  11. Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Oct 2004
    Posts
    121
    Rep Power
    15
    Wow! Thanks! I didn't expect such a positive response on this!

    I was thinking, what if my original site was suffering from some other penalty? If that were the case, wouldn't the links coming out of the site be a lot less effective (if effective at all)? I would think that a link from a penalized site wouldn't boost a site any higher than the penalized site itself.

    And to answer some of your questions, Dazzlindonna:

    Store site:
    * Indexed around Sept. 26, 2004 (according to the wayback machine).
    * Domain was created Sept. 14, 2004 (according to whois).
    * Contains about 425 pages total.
    * Yahoo says it has 1,050 backlinks.
    * Site is currently a PR 3 (was a steady PR 4 before the last update. I kinda let the link building slide on this one for a while... shame on me. )

    Test site:
    * Site was originally indexed Nov. 29, 1999 in the wayback machine. Last record of the site is Mar 9, 2000. The site was re-indexed about 2 months ago (don't know the exact date... shortly after I got the domain).
    * Domain was registered Feb. 16, 2005 (according to whois).
    * Contains only one page.
    * Yahoo says it has 867 backlinks (I can't figure this one out. The only links to this site are from my store. Could it be counting the link twice? Once with the old anchor text and once with the new? It's still coming up #1 for the non-competitive phrase).
    * Site is currently a PR 3 (I found this surprising... considering the store is also a PR 3).

    By the way. I'm checking the page rank from Firefox... not the G Toolbar. The patent made me a little nervous with the "User Maintained / Generated Data".

    One other thing. The site isn't setting the world on fire in the SERPs. It is around #220 in the SERPs. But the store isn't showing up at all.
  12. #7
  13. Contributing User
    SEO Chat Adventurer (500 - 999 posts)

    Join Date
    May 2004
    Posts
    752
    Rep Power
    15
    Nice finding, worth to do same, I have few abandoned sites.
  14. #8
  15. Contributing User
    SEO Chat High Scholar (3500 - 3999 posts)

    Join Date
    Mar 2003
    Location
    Louisiana, USA
    Posts
    3,874
    Rep Power
    24
    Interesting stuff. You are at the 7 month mark on the store site. That tends to be a "get out of the sandbox" time period for many (but not all). Will be interesting to see if the store site appears within the next 1-2 months.
  16. #9
  17. Contributing User
    SEO Chat Adventurer (500 - 999 posts)

    Join Date
    Apr 2004
    Posts
    573
    Rep Power
    16
    * Domain was registered Feb. 16, 2005 (according to whois).
    Test domain expired and You registered it again ?
    Pharmacea.org - New OP Affiliate Program. Join us!. Generic and Brand medications. High 28% commission.
  18. #10
  19. Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Oct 2004
    Posts
    121
    Rep Power
    15
    Originally Posted by eclipse
    Test domain expired and You registered it again ?
    Back when it was first indexed in 1999, it wasn't my site. It belonged to someone else. They abandoned the domain in 2000. I got this domain because it had a history in the wayback machine. I was hoping that meant that the site was previously indexed by Google (several years ago), and that I could create a new site that would bypass the sandbox. I haven't got everything together to make the new site yet, so I figured I'd use it for testing for now.
  20. #11
  21. No Profile Picture
    Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Apr 2005
    Posts
    221
    Rep Power
    15
    age > *

    in googles eyes.
  22. #12
  23. Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Dec 2004
    Location
    Northern NJ, USA
    Posts
    322
    Rep Power
    15
    Unless I'm totally missing the point here, this would imply that anchor text from a sandboxed site is likely to be as powerful as anchor text from a non-sandboxed site.

    Am I missing the point?

    /*tom*/
  24. #13
  25. Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Oct 2004
    Posts
    121
    Rep Power
    15
    Originally Posted by longcall911
    Unless I'm totally missing the point here, this would imply that anchor text from a sandboxed site is likely to be as powerful as anchor text from a non-sandboxed site.

    Am I missing the point?

    /*tom*/
    That's one of several points, actually. The biggest thing about this test is this:

    One site is seven months old along with the domain. The other site is only two months old, but the domain is several years old (according to the wayback machine). Let's call the 7 month old site/domain site A and the 2 month old site (but 5 year old domain) site B. Site A has over 400 pages and was optimized for the site name as well as a few other phrases. Site B is one page (one sentence, actually) and only has back links from site A alone.

    Here is what is driving me crazy. If site B's only back links are from site A, I would think that site A > site B. But apparently, that's not the case. Not only that, but site A has some specific links pointing to it with a three keyword phrase that it's not ranking well for at all. Site B only has links from site A with only two of the three keywords, but it still showing up better than site A for the three keyword set... simply because the three keyword set is on the page on site B (in the title, header, and my one sentence).

    That's what I can't figure out. Site A is better optimized for the keywords and is giving site B its only back links. But site B is doing better. I'm thinking it is the first index date in Google (site A in 09/04 and site B in 11/99), but I guess it could be something else.

    Oh. And one more thing that I think might have been confusing about my original post. I said that the test site says "We've moved!" Just letting you all know that I just put that there to direct people to my store. Honestly, the store was never at that domain. It was the best thing I could think of for throwing up a site in which its only purpose is to send traffic to my store.
  26. #14
  27. Mostly sane...
    SEO Chat Adventurer (500 - 999 posts)

    Join Date
    Aug 2004
    Location
    St. Petersburg, FL
    Posts
    756
    Rep Power
    16
    Olaf,

    Here's a suggestion: remove the link from A to B now that B is indexed. Add pages to B, links from B to A (link them in the subnavigation as if it was one site), and get some decent links to B. See if you can drive traffic by increasing B's placement, at least until A comes out of the sandbox.
    It was a cool test, now you should put the results to use.
  28. #15
  29. Contributing User
    SEO Chat Explorer (0 - 99 posts)

    Join Date
    Mar 2003
    Location
    Weldon, California, USA
    Posts
    59
    Rep Power
    17
    ::: Just Like "Johnny Five" We Need More Input :::

    Okay let's see if we understand your situation correctly.

    Olaf, let us know if we are correct and/or correct us if we are wrong because it appears that we need more pieces to this puzzle to come up with a deductive conclusion. Thanks!


    First of all, we believe that Google ranks PAGES and not SITES. (We all need to really focus in on semantics when it comes to SEO, wouldn't you agree?)

    Therefore, we'll reword what we believe you said. So here goes.


    1.) Domain-A = Your goof-off testing domain.


    2.) Domain-A = WAS indexed in Google several years ago, but was abandoned.

    Our Questions:

    2a.) By "abandoned" do you mean expired?

    2b.) At the time you registered this expired domain were there any pages indexed in Google?

    2c.) If so, about how many?

    2d.) What was the homepage's PageRank if it even had any?

    2e.) How many backward links did Google display for the homepage?

    2f.) How many backward links did Yahoo display for the homepage?

    2g.) How many backward links did MSN display for the homepage?

    2c.) What's the date you registered Domain-A?


    3.) Domain-A = It has been indexed in Google for a couple of months and has been #1 for a VERY non-competitve term.

    3a.) What did you do to index Domain-A in Google? (We assume that you mean the homepage of Domain-A has been indexed when you said ... "It" has been indexed in Google for a couple of months ... Right?) In other words, when you say Domain-A you mean ... < http://www.domain-a.com/ > right?

    3b.) How long did it actually take from the time you started to index Domain-A to the time it was actually indexed by Google.

    3c.) Is it #1 for the term because it has no backward links to it with that term as the anchor text?

    3d.) Is it #1 for the term because it has 1 or more backward links to it (even if the backward link(s) is an intra-domain link)? (DEFINITION: Intra-domain link = A link from one page to another page on the same domain.)


    4.) Domain-A = Anybody can come up for a very non-competitive term. Sandbox or not. I have no problem believing this. The fact is, I changed what keywords I'm targeting for this site now... and it's been in Google long enough to where the "new site bonus" in the SERPs should be gone by now.

    4a.) How long do you believe that the "new site bonus" lasts?

    4b.) Domain-A has been in Google for how long now?


    5.) Domain-B = My existing site (which I believe is sandboxed) is a store. This site has hundreds of pages. Let's call this site "Blue Widgets Store".


    6.) Domain-B = A lot (but not all) of my back links has "Blue Widgets Store" as the anchor text.

    6a.) When you say "a lot" do you mean like 99% have "Blue Widgets Store" as the anchor text?

    6b.) When you say "a lot" do you mean like 95% have "Blue Widgets Store" as the anchor text?

    6c.) When you say "a lot" do you mean like 90% have "Blue Widgets Store" as the anchor text?

    6d.) When you say "a lot" do you mean like 51% have "Blue Widgets Store" as the anchor text?

    6e.) When you say "a lot" do you mean that Domain-B has like 10,000 backward links?

    6f.) When you say "a lot" do you mean that Domain-B has like 1000 backward links?

    6g.) When you say "a lot" do you mean that Domain-B has like 100 backward links?

    6h.) When you say ... has "Blue Widgets Store" as the anchor text ... do you mean that all of these backward links have just that exact phrase or do some or many have additional keywords either before or after your ... "Blue Widgets Store" ... phrase? Do any of your backward links use phrases like ... "Store Blue Widgets?"


    7.) Domain-B = If I do a search for "Blue Widgets Store", my site doesn't show up at all... but it shows up #1 in the allin searches.

    7a.) By "doesn't show up at all" do you mean that your homepage < http://www.domain-b.com/ > doesn't show up in the top 1000 results for the term ... "Blue Widgets Store" ... WITH the quotation marks or doe you mean WITHOUT the quotation marks? (DEFINITION: Placing quotation marks around a term will search for the term as an EXACT PHRASE.)

    7a.) Is it #1 for the < allinanchor: > query like this? ... allinanchor:"Blue Widgets Store"

    7b.) Is it #1 for the < allinanchor: > query like this? ... allinanchor: Blue Widgets Store

    7c.) Is it #1 for the < allinanchor: > query like this? ... allinanchor: Blue "Widgets Store"

    7d.) Is it #1 for the < allintitle: > query like this? ... allintitle:"Blue Widgets Store"


    8.) Domain-A = Now... I've changed my test site to have "Blue Widgets Store" as the title. I've also put it in an H1 tag and in the alt text for an image. The only other thing on this page is a sentence saying "We've moved! Come visit The Blue Widgets Store at www.bluewidgetsstore.com".

    8a.) When you say www.bluewidgetsstore.com you mean ... Domain-B's homepage < http://www.domain-b.com/ > right?

    8b.) Domain-B has how many total pages?

    8c.) Does Domain-B have any intra-domain pages that point to the homepage < http://www.domain-b.com/ >?

    8c.) If no, then does Google still have any pages indexed? They can be seen via their cached pages.


    9.) Domain-B = I then put up a sitewide link from my sandboxed site ...

    9a.) How many total pages contain this link?

    9b.) Did you check to see if Google has "ALL" of these pages indexed?

    9c.) Did all of the pages indexed by Google contain the new updated anchor text?


    10.) Domain-A = ... To this site with "Blue Widgets" as the anchor text.

    11.) Domain-A = About three days later the test site is already in the SERPs for "Blue Widgets Store"... beating my original site that is still nowhere to be found.

    11a.) When you say ... "Blue Widgets Store" ... you really mean ... Blue Widgets Store ... right? In other words, the query you used was WITHOUT the quotation marks, right?

    11b.) If no, then your site wide links from Domain-B will not be factored in. The only factors for your page < Domain-A > being returned in the SERP (Search Engine Results Pages) for the query ... "Blue Widgets Store" ... would be because that EXACT PHRASE is found in the title element <TITLE></TITLE> and in the heading element <H1></H1>.

    11c.) Since you did not mention that your image was also a link, we are assuming that it is not a link. Therefore, the alt attribute text will not be factored into the algorithm.

    11d.) If you answered yes, to question < 11a.) > then did you check the SERP for the query < allinanchor:"Blue Widgets Store" > to see where Domain-A ranks?

    11e.) In question < 7a.) > we asked you if < Domain-B > was #1 for the < allinanchor: > query < allinanchor:"Blue Widgets Store" > and we will now assume that your answer is yes.

    11f.) So, query < allinanchor:"Blue Widgets Store" > right now and let us know where both pages are currently ranked?

    11g.) If we have understood everything that you have said so far, then we will say that < Domain-B > is ranked #1 and that < Domain-A > is NOT ranked at all, nor will < Domain-A > ever be ranked for this query because there are no backward links with the EXACT PHRASE ... Blue Widgets Store ... as its anchor text.


    12.) What kills me is... this site doesn't even have the word "Store" in any of the backlinks, and it's only being supported by a single sandboxed site! It's ranking better than its sole supporter for a keyword phrase the supporting site is aiming for! The supportee is only targeting a keyword subset.

    12a.) Olaf, while your assumption "seems" to be correct based upon the information that you have just shared, it may not be because as you have just read, you didn't provide enough of the pieces to the puzzle for us to be able to completely agree with your assumption at this time. Now, we are not saying that we don't agree with your assumption but only that we need more pieces of the puzzle to be known before we can agree or disagree with your assumption.


    13.) Isn't this kinda odd? Could it be because this test site was once indexed by Google years ago? I don't know what to think. I'm still an SEO newbie, so feel free to flame away.

    13a.) Olaf, anyone who would flame this FANTASTIC post of yours would deserve to have their head examined! Your post is the epitome of peer-based knowledge sharing if you ask us, so we stand and applaud you for it. Kudos Olaf!

    Please do share with us all as much as you can so that we may all learn more. We conduct various types of algorithm analysis and tests and have been doing so since 1996 and when we come across posts like yours we become quite eager to interact, just as you have seen with this post.

    Please understand that we'd love to know more about your situation and to interact with you publicly as much as is possible but please don't share more than you feel comfortable with sharing because we want you to know that we respect your privacy and the need to keep certain things under wraps.

    Olaf, your post will certainly gain the special attention that it deserves, but again, our motive is to interact with you only in such a way that you feel completely comfortable. Therefore, please don't feel that you have to answer each and every question we are asking of you. Answer only those that you feel comfortable answering and no one will think any less of you for it.

    Olaf, thanks again for your most enlightening revelation.

    Your Friends,

    Sharon and Roy Montero


    PS Olaf, please note that we wrote our post right after reading your first and only post in this thread. We now see that you have answered some of our questions so there is no need to repeat your answers. Thanks!

    Comments on this post

    • dazzlindonna agrees : good to see ya'll again. have missed your insights.
    Last edited by Sharon & Roy; Apr 30th, 2005 at 01:31 PM. Reason: To make Olaf aware that we posted using only his initial post.
Page 1 of 2 12 Last
  • Jump to page:

Similar Threads

  1. Skipping the sandbox and a note on sitewides
    By jontelofot in forum Google Optimization
    Replies: 37
    Last Post: Dec 22nd, 2004, 06:49 AM
  2. anyone removed sitewides and got better SERPS?
    By walkman in forum Google Optimization
    Replies: 6
    Last Post: Dec 5th, 2004, 04:56 AM
  3. SERPs change sitewides 'not counted'?
    By cityneil in forum Google Optimization
    Replies: 5
    Last Post: Nov 30th, 2004, 12:08 PM

IMN logo majestic logo threadwatch logo seochat tools logo